Logistic Regression Method for Sentiment Analysis Application on Google Playstore

Abstract

Now days, there are a lot of social media. Each of social media has their one purpose. Each of application has their own function. We can download social media application from Playstore like Google Play or Appstore. Each Playstore gives users the opportunity to provide rating and review. Ratings and reviews from users can give conclusions about the application. Applications with a rating of 5 usually have good comments so that they are more worthy to be downloaded and used. Ratings with a small value cause other users not to use the application. Application owners need to know the results of comments from users which sometimes do not match the value given in the rating. So that the owner or application developer conducts an analysis of user comments on various social media, one of which is on the Playstore. Social media analysis is one way to get public opinion on a topic. One of them is used to analyze various applications provided in the Playstore. Various methods are used to perform social media analysis. This study uses the logistic regression method to analyze public comments into positive or negative comments. We analyze comments from social media applications on Google Play for android users. We get an accuracy value of 81% from 4 social media applications, with a total of 2268 comment data.